Granite 4.2 3B vs Qwen3.8 27B (Xhigh)
Which is better for the money: Granite 4.2 3B by IBM or Qwen3.8 27B (Xhigh) by Alibaba? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-26.
Trade-off
Verdict
Depends on your budget
Qwen3.8 27B (Xhigh) scores 24.6 points higher (33.7 vs 9.1) but costs 98× as much to run the benchmark ($1459 vs $14.9; premium vs budget cost classes). They are not alternatives at one budget: pick Qwen3.8 27B (Xhigh) when you need intelligence above 9.1, and Granite 4.2 3B when 9.1 is enough.
| Metric | Granite 4.2 3B | Qwen3.8 27B (Xhigh) | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 9.1 | 33.7 | +270.3% Qwen3.8 27B (Xhigh) is 3.7× more intelligent |
| Total AA cost (lower is better) | $14.91 | $1,458.94 | +9685.0% Qwen3.8 27B (Xhigh) is 98× more expensive |
| End-to-end time (lower is better) | 63.70s | 61.13s | −4.0% Qwen3.8 27B (Xhigh) is 4.0% faster |
| Price per 1M tokens (in / out) | $0.03 / $0.12 | $0.50 / $3.00 | |
| Input modalities | Text | Text, Image, Video |
Qwen3.8 27B (Xhigh) vs Granite 4.2 3B: 270.3% more intelligence, 9685.0% more cost to run the AA benchmark, and 4.0% faster end-to-end.
Today's Pareto frontier
Against the frontier
- Granite 4.2 3B is off the frontier: at the same or lower cost, GPT-6 Luna (Low) scores 21.50 for $10.6.
- Qwen3.8 27B (Xhigh) is off the frontier: at the same or lower cost, GPT-6.1 Sol (Max) scores 51.80 for $1082.
Granite 4.2 3BQwen3.8 27B (Xhigh)Today's Pareto frontierOpen-weights frontierEvery other model
Matchup
Compare something else
Agents:
Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/granite-4-2-3b/qwen3-8-27b · Markdown: /api/v1/compare/granite-4-2-3b/qwen3-8-27b?format=md · llms.txt · agent skill